Altogether, 69 out of 98 ®bromyalgia (FMS) patients who attended a 2-week multidisciplinary inpatient course and a 1-week control period 3 months later completed 2-km walking tests and stepwise-increased cycling tests at entry and after 3 months. The purpose was to compare the feasibility of the two ®tness tests for assessment of FMS patients. The results showed a substantially lower ®tness level in the results of the 2-km walking test than in the cycling test in the same patients: mean maximum VO 2 was 28.5 ml/kg per min vs 34.6 ml/kg per min, respectively. At entry and after the 3-month training period, the correlations between the two tests were very poor, i.e., with Spearman's r coecients of 0.37 and 0.34, respectively (P<0.01), intraclass correlation coecients (ICC) 0.20 (95% CI ±0.29 to 0.50) and 0.47 (95% CI 0.15 to 0.67), reliability coecients (alpha) 0.54 and 0.47, and Kendal-T coecients 0.32 and 0.41 for ordinal correlation of the test results. The results did not correlate with pain, which remained at initial levels for the 3 months of follow-up. Principally, these tests should both measure the same property, i.e., the ®tness of ®bromyalgia syndrome (FMS) patients, but the results diered substantially. The 2-km walking test showed a markedly lower ®tness level than the cycling test in the same patients. The primary explanation for this dierence might be diculties in controlling test performance, especially in walking. The 2-km walking test would not appear recommendable for subjects with chronic pain syndrome, e.g., ®bromyalgia.